Buffalo Forums
Products => Storage => : cr1m February 07, 2019, 04:23:45 AM
-
Hello, we using the storage server TS-RXLAFE. Two days ago it was updated to 1.72 version software.
And we have that problem:
If we want to copy any files to open share or on local system on this server, we see message:
root@TS-RXLAFE:/mnt/array1# cd backup/
root@TS-RXLAFE:/mnt/array1/backup# mkdir test
mkdir: cannot create directory 'test': No space left on device
But we can read files and delete it. We can't only copy and create on mapped drive.
Mapped drive is xfs system (raid 5 with 8Tb volume)
root@TS-RXLAFE:/mnt/array1/backup# df -h
Filesystem Size Used Available Use% Mounted on
/dev/md1 4.7G 1.0G 3.4G 23% /
/dev/ram1 15.0M 204.0k 14.8M 1% /mnt/ram
/dev/md0 969.2M 36.7M 932.5M 4% /boot
/dev/md2 8.1T 6.1T 2.1T 74% /mnt/array1 -- 2.1Tb free!
Many free inodes we have
root@TS-RXLAFE:/mnt/array1/backup# stat -f /mnt/array1
File: "/mnt/array1"
ID: 902 Namelen: 255 Type: xfs
Block size: 4096
Blocks: Total: 2185632256 Free: 558223994 Available: 558223994
Inodes: Total: 1748923136 Free: 1748828992
Version of linux
root@TS-RXLAFE:/mnt/array1/backup# uname -a
Linux TS-RXLAFE 2.6.31.8-mv78100 #2 Mon Dec 18 12:35:25 JST 2017 armv5tel unknown
Can anybody help us? Why its happening?
-
Are you using quotas?
-
Are you using quotas?
/dev/md2 on /mnt/array1 type xfs (rw,relatime,attr2,nobarrier,sunit=1024,swidth=3072,noquota)
noquota.
And I can't create files and dirs in system (without network)
-
User quota on /mnt/array1 (/dev/md2)
Blocks Inodes
User ID Used Soft Hard Warn/Grace Used Soft Hard Warn/Grace
---------- --------------------------------- ---------------------------------
root 8K 0 0 00 [------] 11 0 0 00 [------]
nobody 6.1T 0 0 00 [------] 94.1k 0 0 00 [------]
Group quota on /mnt/array1 (/dev/md2)
Blocks Inodes
Group ID Used Soft Hard Warn/Grace Used Soft Hard Warn/Grace
---------- --------------------------------- ---------------------------------
root 8K 0 0 00 [------] 11 0 0 00 [------]
nogroup 6.1T 0 0 00 [------] 94.1k 0 0 00 [------]
root@TS-RXLAFE:~# xfs_quota -x -c "disable -gu -v" /mnt/array1
User quota state on /mnt/array1 (/dev/md2)
Accounting: ON
Enforcement: OFF
Inode: #2051 (3 blocks, 3 extents)
Group quota state on /mnt/array1 (/dev/md2)
Accounting: ON
Enforcement: OFF
Inode: #2052 (2 blocks, 2 extents)
Blocks grace time: [7 days 00:00:30]
Inodes grace time: [7 days 00:00:30]
Realtime Blocks grace time: [7 days 00:00:30]
Doesnt helped.
-
I'm kind of hoping someone who's dealt with something like this on these devices will chime in, I've run into errors that resemble this in the past but they were all scenarios that I doubt would apply in this case.
one thing you probably want to try is running an xfs_check and see if the filesystem reports as healthy, I beleive certain types of filesystem issue manifest as filefull rather than io errors but I don't know any details.
specifically, my first drive with 4k sectors did not play well with the motherboard I had at the time which resulted in it reporting 3TB of space but in practice it could only access 750GB (I think). Sadly i didn't notice until I started getting weird behavior when it tried to write past that part of the drive.
The other was when I tried creating an 18TB raid array on a 32-bit arm device (lsqvl). Similarly it would let me create the volume and even format it but any attempts to access blocks past the 16TB mark would fail.
I wouldn't think either of these would apply to your situation, particularly if these are the original disks that came with the device. But if it's happening when you try to write things from the OS as root and not just over samba, that narrows it down a bit at least.
-
I did check with xfs_repair -p, it doesnt found any problems. Then I started the rebuilding raid, but now 2 days ago it still in 5% of rebuild.
-
you can reflash an older firmware (read about debug options from TSUpdater)
Are all drives ok?
Read out the smart status of any drive
-
Can U help me with searching for older firmware? On official site I see only 1.72 version.
-
I might have missed a post, why older firmware?
-
Two days ago it was updated to 1.72 version software.
And we have that problem:
-
ah, ok.
-
1.66
https://drivers.softpedia.com/get/HDD-SSD-NAS-USB-Flash/Buffalo/Buffalo-TS-RXL-R5-NAS-Firmware-166.shtml (https://drivers.softpedia.com/get/HDD-SSD-NAS-USB-Flash/Buffalo/Buffalo-TS-RXL-R5-NAS-Firmware-166.shtml)
-
It's probably worth pointing out that the past few firmware updates have been to patch some pretty scary samba vulnerabilities. Probably worth a shot while troubleshooting but I wouldn't recommend leaving it running that version long term.
-
try
df -i
from
https://askubuntu.com/questions/590315/no-more-space-available-even-if-there-is-a-lot (https://askubuntu.com/questions/590315/no-more-space-available-even-if-there-is-a-lot)
-
Inodes: Total: 1748923136 Free: 1748828992
I think he covered that in the original post
-
Problem was solved with rebuilding raid array!
Thanks for all!